How social are social media? A cross-cultural comparison of online and offline purchase decision influences

Transcription

1 Journal of Marketing Communications, How social are social media? A cross-cultural comparison of online and offline purchase decision influences Kendall Goodrich a * and Marieke de Mooij b a Department of Marketing, Raj Soin College of Business, Wright State University, 279 Rike Hall, 3640 Colonel Glenn Highway, Dayton, OH 45435, USA; b Cross-Cultural Communications Consultant, Westerenban 44, 4328 HE Burgh-Haamstede, The Netherlands Traditional word of mouth (WOM) involves personal communications among family, friends, and others. Today, social media sites like Facebook offer the possibility of electronic word of mouth (ewom), a mediated form of WOM. Little is known about the best way for marketers to use social media. Even less is known about international differences in consumer usage of social media and how ewom influences consumer decision-making. This study helps address that research gap and uses cultural dimensions to compare the use of social media and other information sources for consumer decision-making across 50 countries. The results indicate that the use of information sources that influence online purchase decisions strongly varies by culture. For example, relationship-oriented collectivists rely to a greater extent than individualists on social media, which are an alternative for interpersonal WOM communication. Interestingly culture affects the influence of social media on purchases differently than the influence of traditional WOM through family or friends. Furthermore, there are major differences in online complaint behavior by country, due to cultural variations. Keywords: social media; online; purchase; culture; individualism; collectivism; word of mouth Introduction Online social networking is growing rapidly worldwide. Naturally, marketers want to reach potential customers with this new channel. Already in 2011, 25% of the Fortune Global 100 were using all major social media platforms to increase social media presence, whereas 84% were on at least one platform (comscore 2011). However, little research is available about the function of social media and other online marketing communication tools, particularly regarding their international applicability. Social media are hybrid media in the sense of offering both electronic word of mouth (ewom) opportunities and serving as a mass media platform for sending messages. WOM, or informal communications directed at other consumers about the ownership, usage, or characteristics of particular goods and services and/or their sellers (Westbrook 1987, 261), has typically been viewed as an effective influence on purchase decisions. Now, ewom can be directed from one individual to many consumers online. The purpose of this study was to examine the influence of culture on the role of social media in the consumer decision-making process as compared with other online or offline sources of information. The study contributes to the marketing literature by examining the influence of culture on the use of online and offline purchase decision information sources across 50 countries. *Corresponding author. kendall.goodrich@wright.edu q 2013 Taylor & Francis

2 2 K. Goodrich and M. de Mooij Social networking and culture Globalization has not resulted in converging consumer behavior (De Mooij 2011). People around the world still need to feel culturally engaged with vendors (Lynch and Beck 2001) or else they will find another place to buy. Also the Internet is not leading to such convergence. For example, cultural differences across countries with respect to products and services bought on the Internet appear to mirror differences found in traditional shopping channels (Goodrich and De Mooij 2011). The way people use the Internet varies worldwide. This applies to usage frequency, number and type of contacts, interactivity, and content. For example, the topics dominant in blogs or social media in Asia are different from those in the West. Results from a study across 22 countries (Su et al. 2005) comprising southeast Asia, Japan, North America, and Europe suggest the differences in topics covered as well as blogging motivations. Japanese blogs, for example, are dominated by hobby and recreational topics, whereas personal blogs dominate in most other cultures. The number of friends reported by users of social media also varies greatly from 29 in Japan, 63 in China, 95 in France, 200 in the USA to 360 in Brazil (Van Belleghem 2010). Several comparative studies have found that worldwide social media usage focuses on contacts with friends and family. For example, a comparison of Hispanic students in Argentina, Chile, Colombia, the USA, and Uruguay showed that the main use of social media was to maintain contact with friends and family. Most liked activities were keeping in touch, sharing photos and reconnecting with old friends (Albarran and Hutton 2010). However, differences were also found. A comparison of social media behavior across five other countries (the USA, the UK, Italy, Greece, and France) found that social browsing was more important for French and Italian users than for US users. For the French users, status updates and photographs were less important than for the US users (Vasalou, Joinson, and Courvoisier 2010). To much of the English-speaking world, it might appear that Facebook has become ubiquitous, but in many countries, Facebook has relatively low penetration. Local social networking sites generate more traffic than Facebook, for example, in China (renren.com), in Japan (mixi), in South Korea (Cyworld), and in Russia (vkontakte.ru). The percentage of Facebook users among total online users is 68 in the USA and 62 in the UK, but only 14 in Japan, 22 in Korea, 10 in Russia, and,1 in China (CheckFacebook.com 2012). The success of local network services is due to their cultural appropriateness. An important difference among cultures might also be social media s function in the consumer decision-making process. An important question is whether social media are viewed as sources of information, similar to the traditional media, or as influences in opinion formation, similar to interpersonal or WOM communication. To understand how culture influences the way people communicate and use media, dimensional models can be used that delineate national cultural values and help explain and predict behavior (Hofstede 2001). Hofstede s cultural variables Hofstede identified five factors that contribute to the understanding of national cultural values and explain differences in human behavior, labeled (1) individualism (IDV)/collectivism (COL); (2) uncertainty avoidance; (3) power distance; (4) masculinity/ femininity; and (5) long-/short-term orientation. The dimensions are measured on a scale from 0 to 100 for 75 countries, and each country has a position on each scale or index. These country scores can be used for statistical analysis of data on consumer attitudes and

3 Journal of Marketing Communications 3 behavior. Several models have been developed in the past decades, but Hofstede s model is used most widely in marketing studies (Magnusson et al. 2008; Soares, Farhangmehr, and Shoham 2007). Some critics doubt the validity of the Hofstede model, because the original measurements were done within IBM corporation. However, this methodology allowed comparison of matched samples to discover differences caused solely by national culture, not by different corporate cultures. Since then, the study has been frequently replicated on different populations (Søndergaard 1994). In the second edition of his book Culture s Consequences (2001), Hofstede describes over 200 external comparative studies and replications that have supported his indexes. A reason for widespread adoption of Hofstede s classification of culture lies in the large number of countries measured and the simplicity of dimensions, which are straightforward and appealing to both academic researchers and business people. A comparison of models for international marketing strategy (Magnusson et al. 2008) shows that more recent cultural frameworks provide little advancement compared with Hofstede s work. Many data on product ownership and related behavior (De Mooij 2010, 2011) appear to correlate with Hofstede s dimensions. Therefore, Hofstede s cultural variables are used in this study, and we formulate hypotheses about how these dimensions explain differences in the influence of the various purchase decision sources across cultures. Four of the five dimensions that have been shown to explain differences in communication behavior are used for this study. The following summary provides greater depth into these four dimensions, including specific prior research applications. Power distance (PDI 1 /2) PDI reflects the extent to which less powerful members of a society accept and expect unequal power distribution. In high PDI cultures, everyone has a rightful place in the social hierarchy, but in low PDI cultures, equality and independence are highly valued (Hofstede, Hofstede, and Minkov 2010). Anglo-Saxon and northwest European countries score lowest on PDI with much of the rest of the world scoring higher. In low PDI cultures, people depend less on other people and base decision-making more on facts and data, consciously gathering information throughout the decision-making process. They read more newspapers and watch less television than in high PDI cultures (De Mooij 2011). Consumers online research into brands is stronger in cultures of low PDI (data from Mediascope Europe 2008). In high PDI cultures, people rely more on personal sources of recommendation, are more active opinion seekers, and are less active in information seeking via impersonal sources (Dawar, Parker, and Price 1996; Pornpitakpan 2004). Low PDI Western cultures are more likely to use websites emphasizing marketer-to-consumer interactivity than high PDI Eastern cultures, where individuals favor consumer-to-consumer interactivity, due to a larger distance between marketers and consumers (Cho and Cheon 2005). Individualism/collectivism IDV (vs. COL) can be defined as people looking after themselves and their immediate family only versus people belonging to in-groups who look after each other in exchange for loyalty. In individualistic cultures, one s identity is in the person. In collectivistic cultures, people are we -conscious. Their identity is based on the social system to which they belong, and upholding face is important.

4 4 K. Goodrich and M. de Mooij In individualistic cultures, individuals search for information to maximize their personal utility, whereas in collectivistic cultures, individual utility is less important than sharing with others, so that new electronic media are used more for sharing ideas and opinions than for personal information search. Paradoxically, family and friends are viewed as more important in individualistic cultures than in collectivistic cultures (Eurobarometer 2007), where family and friends are automatically part of one s identity. In individualistic cultures, friendship is explicitly valued because people have to make an effort to make friends and preserve friendship. In collectivistic cultures, people meet others more frequently and there is more interpersonal communication, which extends to WOM about products and brands. In China, the major influence on purchase decisions is WOM (Schultz and Block 2009). The influence of WOM is intensified by the various discussion opportunities offered by the Internet. Members of collectivistic and of high PDI cultures tend to share ideas and opinions orally, but also chat more often than members of individualistic and low PDI cultures (De Mooij 2011). In collectivistic cultures, the importance of harmony and maintaining face cause people who experience post-purchase problems to avoid voicing complaints directly to the provider. For example, the Chinese are less likely to lodge a formal complaint about a faulty product than Australians (Lowe, Chun-Tung, and Corkindale 1998), but they do engage in negative WOM to in-group members. About 70% of the world population score collectivistic only the Anglo-Saxon cultures, the Netherlands, Belgium, France, Germany, and the Scandinavian countries score individualistic. The IDV/COL dimension is particularly important for understanding differences in online buying influences because it explains differences in personal or nonpersonal communication behavior. Worldwide IDV is correlated with low PDI, so when correlations are found with the one dimension, they are often found with the other. Uncertainty avoidance (UAI 1 /2) UAI is the extent to which people feel threatened by and try to avoid or reduce uncertainty and ambiguity. It is not the same as risk avoidance. According to Hofstede, Hofstede, and Minkov (2010, 197), UAI cultures shun ambiguous situations and paradoxically they may be prepared to engage in risky behavior to reduce ambiguities. Generally, members of high UAI cultures have low trust in people and institutions. For example, agreement with the statement most people can be trusted (World Values Survey 2011) correlates significantly with low UAI. In low UAI cultures, more opinion seekers are found (Pornpitakpan 2004), but people also search more heavily for information from impersonal sources (Dawar, Parker, and Price 1996). In low UAI cultures, therefore, consumers typically base their decisions on more information sources than do consumers in high UAI cultures, where feelings of trust dominate decision-making. In Europe, countries in the Northwest score low on UAI and countries in Southcentral and eastern Europe score high. All of Latin America scores high on the UAI index and Asia shows a mix; China, for example, scores low, but Japan and Korea score high. Long-/short-term orientation Long- versus short-term orientation is the extent to which a society exhibits a pragmatic, future-oriented perspective, rather than a short-term point of view. Originally, this

5 Journal of Marketing Communications 5 dimension was measured for a limited number of countries, but recent new measurements (Hofstede, Hofstede, and Minkov 2010) provide data for nearly all countries which have scores on the other dimensions. This dimension has not been used much because the data are relatively new, so few earlier studies are available. Included in short-term orientation are values of national pride, tradition, low thrift, importance of service to others, and selfenhancement. Included in long-term orientation are thrift, perseverance, and pragmatism. This dimension differentiates collectivistic cultures with respect to communication style and how people present themselves. Collectivistic cultures with short-term orientation in Latin America communicate in a more expressive and elaborative way and prefer to enhance the self in their presentation, whereas collectivistic cultures of long-term orientation in Asia show a more succinct style and prefer a more modest self-presentation. The need for self-enhancement results in a preference for social media that facilitate strong self-presentation. In particular, American-style social media promote selfenhancement. So it is not surprising that the percentage of Facebook members (CheckFacebook.com 2012) as well as the numbers of friends reported across cultures correlates with short-term orientation. Brazil scores low on the long-term orientation scale, which explains the tendency for Brazilians to be very image conscious and carefully craft their online persona. Image and reputation are important parts of status and guide online behavior (Specht 2010). Long-term orientation also explains differences in literacy (Adult literacy rates UN development program 2009) and newspaper circulation (Unesco Institute for Statistics 2011) with higher literacy rates and more newspapers in long-term-oriented cultures. East Asian countries score high on long-term orientation, whereas the Anglo-Saxon countries, Latin American, and African countries score low. European countries vary with Estonia scoring highest and Portugal lowest. Opinion and information seeking in the consumer buying process IDV/COL and PDI best explain differences in the way people acquire information and the importance of information in the decision-making process, in particular with respect to active search for information. In individualistic cultures, information is an allencompassing need, whereas in collectivistic and high PDI cultures, trustworthiness and the opinions of others are more important. People are less active information seekers, acquire information passively through many personal contacts, and base their buying decisions more on feelings and company trust via WOM, rather than on hard facts. In individualistic and low PDI cultures, however, people search for information more actively, whether from WOM or traditional media. Personal contacts serve more as a source of information, whereas in collectivistic and high PDI cultures, personal contacts tend to serve to form opinions. Boase et al. (2006) confirm that social media, along with search engines and other online information sources, are viewed as a path to resources in the Western individualistic world. Social media are used to access people who might have the right information to help solve a problem, whereas in collectivistic and high PDI cultures, social media tend to reinforce the sharing of feelings and ideas. However, not all collectivistic cultures have the same information preferences with differences explained by long-term orientation. One of our hypotheses is that consumers from short-term-oriented cultures rely more on the human factor and prefer to get their information from people, whereas in long-term-oriented cultures, consumers rely more on facts and data. These cultural differences might also apply to various online purchase decision sources.

6 6 K. Goodrich and M. de Mooij Traditional opinion and information sources such as magazines and newspapers, although still important to many, are largely being surpassed worldwide by online versions of magazines and information sources such as search engines, product information sites, and so on. An important research question is whether new media mirror the old media in terms of trust and influence on consumer behavior. Do social media have the same function as interpersonal WOM in the decision-making process? If newspapers had an information function, do search engines, for example, have the same function? If so, we might expect that the cultural variables that explain differences in usage of the old media will also explain variance of the new media. Hypotheses Although many studies compare limited numbers of countries, we were able to compare a larger number of countries to assess the differences across nations with respect to social media and other Internet influences. Our major research questions are (1) whether members of different cultures are attracted to different Internet information and communication applications such as social media, search engines, forums and so on.; (2) whether these differences mirror variations in the usage of traditional media or interpersonal communications and their roles in information seeking or opinion formation; and (3) which cultural variables best explain these differences. Opinion seeking Previous studies have shown that personal sources of information and opinion seeking are more important in higher PDI and collectivistic cultures (Dawar, Parker, and Price 1996; Pornpitakpan 2004), suggesting greater reliance on social network connections and online forums for making decisions. Therefore, usage of social media and trust in online forums should influence online purchase decisions in cultures scoring low on IDV and high on PDI. H1a: Usage of social media for purchase decision-making will be correlated negatively with IDV and positively with PDI. H1b: Trust in online forums for purchase decision-making will be correlated negatively with IDV and positively with PDI Friends are particularly important in individualistic and low PDI cultures. In these cultures, friends opinions should also be trusted for making purchase decisions. H1c: Trust in friends for making purchase decisions will be positively correlated with IDV and negatively with PDI Information seeking People from individualistic cultures have fewer interpersonal contacts, read more than people from collectivistic cultures, and tend to search for information to maximize personal utility. Therefore, members of individualistic cultures are expected to prefer more factual information sources such as search engines, websites, and online product reviews than are people from collectivistic cultures. H2a: Trust in search engines for purchase decision-making will be correlated positively with IDV and negatively with PDI.

7 Journal of Marketing Communications 7 H2b: H2c:: Trust in websites for purchase decision-making will be correlated positively with IDV and negatively with PDI. Trust in online product reviews for purchase decision-making will be correlated positively with IDV and negatively with PDI Human-based or fact-based information Members of low UAI cultures generally have more trust in people, whereas members of high UAI cultures have more trust in experts. Search for nonpersonal facts might be expected to reduce uncertainty and thus relate to high UAI. Therefore, trust in purchaserelated recommendations from informational online media, such as product websites and search engines, are expected to be positively related to UAI, whereas trust in friends might be negatively related. H3a: Trust in recommendations by friends for purchase decision-making will be negatively correlated with UAI. H3b: Trust in product websites for purchase decision-making will be positively correlated with UAI. H3c: Trust in search engines for purchase decision-making will be positively correlated with UAI The long-/short-term orientation dimension differentiates between collectivistic cultures with respect to the importance of fact-based information and human-based information. The relationship between long-term orientation, literacy, and newspaper circulation may point at a higher degree of nonpersonal information seeking in long-termoriented cultures, which also applies to Internet applications such as search engines and product reviews. Short-term-oriented cultures are more likely to trust recommendations from friends and family as well as recommendations by people on television. H4a: Trust in recommendations by family for making purchase decisions will correlate negatively with long-term orientation. H4b: Trust in recommendations by friends for making purchase decisions will correlate negatively with long-term orientation. H4c: Trust in recommendations by people on TV for making purchase decisions will correlate negatively with long-term orientation. H4d: Trust in search engines for making purchase decisions will be positively correlated with long-term orientation. H4e: Trust in online product reviews for making purchase decisions will be positively correlated with long-term orientation Sharing negative information online In collectivistic and high PDI cultures, people who experience post-purchase problems avoid complaining to the supplier (e.g., Lowe, Chun-Tung, and Corkindale 1998) because they are reluctant to make other people lose face. However, they do engage in negative WOM with in-group members. Social media and other online resources provide new channels for sharing negative WOM among the in-group. Thus, we would expect COL to be related to sharing negative experiences online. H5: Sharing negative information online will be negatively correlated with IDV and positively with PDI

8 8 K. Goodrich and M. de Mooij Methodology The dependent variables in this study are results from a survey conducted by Nielsen (Nielsen 2010) on the influence of social media on purchase decisions and trust in recommendations from various online and offline influences. The specific social media question was: I use social media sites to help me make purchase decisions. The trust questions related to recommendations by family, friends, product websites, search engines, online forums, and recommendations from TV ( I trust recommendations from these sources most when making a purchase decision... ). The resulting percentages of people relying on the various sources, by country, were used as dependent variables. When referring to recommendations, this was a general way to refer to influences, whereby recommendations from TV would refer to people and content on TV and recommendations from search engines would include information delivered in response to search requests. The Nielsen survey question concerns the TV medium as such, not specific contents. It would be difficult to specify exact TV content, since this tends to differ worldwide. The Nielsen online survey, which used 37 languages, took about 15 min to complete and examine how consumers shopped online, what they intended to buy, how they used various sites, and the impact of social media and other influences on the purchase decision. Nielsen polled over 27,000 Internet users in 55 countries from all parts of the world Asia Pacific, Europe, Middle East, North America, and South America (see Table 1 for Table 1. Countries surveyed and sample sizes. Argentina 502 Malaysia 502 Australia 501 Mexico 502 Austria 502 Netherlands 498 Belgium 501 New Zealand 500 Brazil 500 Norway 510 Canada 500 Pakistan 500 Chile 500 Philippines 509 China 500 Poland 501 Colombia 500 Portugal 501 Croatia 502 Romania 502 Czech 500 Russia 500 Denmark 500 Saudi Arabia 506 Egypt 455 Singapore 500 Estonia 501 South Africa 500 Finland 501 South Korea 501 France 505 Spain 507 Germany 502 Sweden 503 Greece 502 Switzerland 500 Hong Kong 500 Taiwan 503 Hungary 502 Thailand 532 India 502 Turkey 546 Indonesia 505 UAE 504 Ireland 511 UK 508 Israel 503 Ukraine 509 Italy 501 US 509 Japan 502 Venezuela 502 Latvia 501 Vietnam 504 Lithuania 503 From the Nielsen Global Online Consumer Survey, The Nielsen Company Age and gender quotas were used to produce a sample representative of adult Internet users in each country.

9 Journal of Marketing Communications 9 countries surveyed). Samples consisted of at least 500 respondents per country and were stratified according to age and gender quotas to produce samples that are representative of adult Internet users in each country. The cultural variables of PDI, IDV, UAI, and longterm orientation were used to evaluate cultural differences by country (Hofstede, Hofstede, and Minkov 2010). Country scores for the Hofstede dimensions can be downloaded from Correlation analysis and stepwise linear regression were used to evaluate cultural influences (e.g., De Mooij and Hofstede 2002) on the use of social media and other sources in the purchase decision-making process (see Table 2 for the summary of results). There were eight dependent variables, as shown in Table 2. The independent variables were four of Hofstede s dimensions. For long-term orientation, one country (United Arab Emirates) was missing, but Geert Hofstede provided an estimate. In most cases, national wealth (GNI/capita) did not play an explaining role, so we did not include wealth as an independent variable. Correlation analysis is one tailed. Significant levels are indicated by *p, 0.05, **p, 0.01, and ***p, When regression analysis is used, multiple linear regression analysis has to be carried out stepwise. The coefficient of determination, or R 2, is the indicator of the percentage of variance explained. Results Pearson correlations between both usage of social media data and trust in online forums were significant and positive with PDI (r ¼ 0.71***, r ¼ 0.27*) and negative with IDV (r ¼ ***, r ¼ *), supporting H1a and H1b. In stepwise regression, the combination of low IDV and high PDI together explained 67% of variance in social media Table 2. Sources for making purchase decisions and cultural differences (correlation/regression). Dimensions Pearson r Stepwise regression, significant variables R 2 Use social media sites to help make PDI 0.71*** IDV 2, 0.58 purchase decisions IDV *** PDI þ, 0.67 Trust recommendations from family LTO *** LTO 2, 0.36 Trust recommendations from friends UAI *** UAI 2, 0.22 LTO *** LTO 2, 0.39 Trust recommendations from product websites PDI 0.25* UAI þ, 0.16 IDV * UAI 0.41*** Trust recommendations from search engines UAI 0.35** LTO þ, 0.25 LTO 0.50*** UAI þ, 0.37 Trust recommendations from online forums PDI 0.27* IDV * Trust recommendations from TV PDI 0.37*** LTO 2, 0.28 IDV *** IDV 2, 0.53 LTO *** Share negative experience PDI 0.48*** IDV 2, 0.25 IDV *** r ¼ Pearson product moment correlation coefficient. Correlation analysis is two tailed. Significance levels are indicated by *p, 0.05; **p, 0.01; ***p, Multiple linear regression analysis is done stepwise. R 2 ¼ percentage of variance explained.

10 10 K. Goodrich and M. de Mooij usage. Stepwise regression did not result in explanations for variance in trust in online forums. No significant correlation was found between trust in friends for purchase decision-making and IDV or low PDI, so H1c was not supported. Perhaps the importance of friends does not include using them online for help in purchase decisions. Although friends do play a role in WOM in individualistic cultures, our findings suggest that this may not apply to ewom. No significant correlations were found between IDV and/or low PDI and trust in search engines and product reviews, failing to support H2a and H2c; marginally significant correlations were found with trust in websites, providing only marginal support for H2b. Significant correlations were found between low UAI and recommendations by friends (r ¼ ***) as well as between high UAI and trust in product websites (r ¼ 0.41***) and search engines (r ¼ 0.35**), supporting H3a, H3b, and H3c. Low UAI explained 22% of variance in trust in recommendations by friends and high UAI explained 16% of variance in trust in websites. This supports findings by Pornpitakpan (2004) that people are susceptible to interpersonal influences in low UAI cultures. UAI together with long-term orientation explained 37% of variance in trust in search engines. With respect to the long-/short-term orientation dimension, significant correlations were found between short-term orientation and trust in family (r ¼ ***) and friends (r ¼ ***), as well as between long-term orientation and trust in TV (r ¼ ***). Short-term orientation uniquely explained 36% of variance in trust in family. Together with low UAI short-term orientation explained 39% of trust in friends, and together with COL, it explained 53% of trust in recommendations from TV. Thus, H4a, H4b, and H4c were supported. H4d and H4e were also supported because significant correlations were found between long-term orientation and both trust in search engines (r ¼ 0.50***) and trust in online product reviews (r ¼ 0.41***). Long-term orientation explained 17% of variance in trust in product reviews, and together with high UAI explained 37% of variance in trust in product reviews. H5 was supported, since sharing a negative experience via social media or other online sources was correlated negatively with IDV (r ¼ ***) and positively with PDI (r ¼ 0.48***). High PDI explained 22% of variance. Discussion, implications, and future research The results of this study indicate that Hofstede s cultural dimensions explain cross-cultural differences in both online and offline purchase decision influences. An important finding is the strong explaining function of IDV/COL for the usage of social media across cultures. Marketers should consider using social media more in collectivistic cultures. For this purpose, cultures can be mapped with respect to IDV/COL as shown in Figure 1, illustrating the differences in social media use. The x-axis shows the scores on the IDV COL dimension and the y-axis shows the percentages of social media usage. The UK, the USA, and Australia are examples of individualistic countries, which are less likely to use social media for purchase decisions than are collectivistic cultures like China and Thailand, where social media may play an important role in opinion formation. Although the IDV COL dimension is often used to explain differences in communication behavior, this study demonstrates that other dimensions, e.g., UAI and long-/short-term orientation, also play a role. For example, UAI and long-term orientation explain variations in trust in online sources such as search engines. Our findings also suggest important cross-cultural differences between the role of human beings such as family and friends through WOM or ewom and inanimate sources of information such as

11 Journal of Marketing Communications 11 Figure 1. Individualism/collectivism and usage of social media for purchase decisions. search engines and websites, which function more like some of the traditional media. Preferences for human sources of information are more important in cultures of short-term orientation and low UAI, whereas fact-based information sources such as search engines are more important in cultures of long-term orientation and high UAI. The long-/short-term orientation dimension also differentiates the function of social media across collectivistic cultures. Another useful 2D map (Figure 2) combines COL IDV and short- and long-term orientation. It shows four different quadrants that can be taken into account when trying to generate ewom through social media. Although in short-term-oriented collectivistic cultures, people identify and present themselves in a self-enhancing way and are more interactive, in long-term-oriented collectivistic cultures, people want to be anonymous and feel part of the larger community in a more passive way. This explains why Facebook is more popular in the short-term-oriented cultures, whereas a long-term oriented culture like Japan has developed its own, culturally appropriate social network site. Culture-related motives for social media such as self-enhancement and status are stronger in South America than in East Asia. Marketers therefore should consider using brand messages, which emphasize self-expression to these short-term orientation cultures, but emphasize group orientation and anonymity to long-term orientation countries with additional presence on social media sites other than Facebook. Across individualistic cultures in the two right-hand quadrants long-/short-term orientation explains differences with respect to the degree of self-enhancement, for example, expressed by the difference in numbers of friends. Marketers should also consider using social media more in

12 12 K. Goodrich and M. de Mooij Figure 2. Function of social media by long-term orientation and individualism. collectivistic cultures and using search engine marketing more in individualistic cultures, thereby further differentiating between short- and long-term-oriented cultures. The results also indicate that in collectivistic and high PDI cultures, people are more actively engaged in negative WOM with in-group members online than are those in individualistic and low PDI cultures. These results imply that marketers should recognize that individuals from these cultures are unlikely to freely offer complaints directly to them and that they must actively monitor complaints elsewhere, whether from social media sites or confidential surveys. This study contributes to the international marketing literature by finding important differences in social media usage and in other online and offline purchase decision influences around the world. Different cultures show varying emphasis on different types of online information and develop local variations of social media that cater to particular cultural usage. We hope that the results spur future academic research into specific countries of interest and that practitioners can better target online marketing communications through improved understanding of global cultural differences. Acknowledgments The authors thank the reviewers of this manuscript for their constructive comments.

13 Journal of Marketing Communications 13 Notes on contributors Kendall Goodrich is an associate professor of marketing at Wright State. He received a Ph.D. in marketing from Florida Atlantic University, Master s of Management from Northwestern University (Kellogg School), and Bachelor of Arts degree from the University of Virginia. Kendall has worked in private sector marketing management positions for AT&T and NCR, has participated in a successful IPO with Citrix, and has held executive marketing positions with multiple e-commerce firms. Dr. Goodrich teaches in the areas of Internet marketing, entrepreneurship, and marketing strategy. His research explores areas of consumer behavior such as advertising stimuli effects, cognitive processing, global cultural differences, and Internet commerce. His writing has appeared in publications such as the Journal of Business Research, Journal of Advertising Research, and Psychology & Marketing and in conference proceedings such as American Marketing Association, Academy of Marketing Science, and Association for Consumer Research. Marieke de Mooij, Ph.D. is a consultant in cross-cultural communications. She is a retired profesora asociada at the University of Navarra (Spain) and has taught at various other universities in the world. She is the author of several academic publications on the influence of culture on marketing, advertising, and consumer behavior. She is the author of Global Marketing and Advertising. Understanding Cultural Paradoxes 4th Edition (2013) and Consumer Behavior and Culture. Consequences for Global Marketing and Advertising. 2nd Edition, both published by Sage. A new book, to be published by Springer, is Human and Mediated Communication Around the World. A Comprehensive Review and Analysis. Her website is References Albarran, A. B., and B. Hutton Young Latinos Use of Mobile Phones: A Cross-Cultural Study. Accessed December 10, ography/cellphonestudypaper.pdf Boase, J., J. B. Horrigan, B. Wellman, and L. Rainie The Strength of Internet Ties. Pew Internet & American Life Project. Accessed August 1, PIP_Internet_ties.pdf CheckFacebook.com Accessed September 1, Cho, C. H., and H. J. Cheon Cross-Cultural Comparisons of Interactivity on Corporate Web Sites: The United States, the United Kingdom, Japan, and South Korea. Journal of Advertising 34 (2): comscore U.S. Digital Year in Review. February. Accessed August 1, Dawar, N., P. M. Parker, and L. J. Price A Cross-Cultural Study of Interpersonal Information Exchange. Journal of International Business Studies 27 (3): De Mooij, M., and G. Hofstede Convergence and Divergence in Consumer Behavior: Implications for International Retailing. Journal of Retailing 78 (1): De Mooij, M Global Marketing and Advertising, Understanding Cultural Paradoxes. Thousand Oaks, CA: Sage. De Mooij, M Consumer Behavior and Culture: Consequences for Global Marketing and Advertising. 2nd ed. Thousand Oaks, CA: Sage. Eurobarometer European Social Reality. Special Eurobarometer Report 273. Goodrich, K., and M. De Mooij New Technology Mirrors Old Habits: Online Buying Mirrors Cross-National Variance of Conventional Buying. Journal of International Consumer Marketing 23 (3/4): Hofstede, G Culture s Consequences. 2nd ed. Thousand Oaks, CA: Sage. Hofstede, G., G. J. Hofstede, and M. Minkov Cultures and Organizations: Software of the Mind. 3rd ed. New York: McGraw-Hill. Lynch, P. D., and J. C. Beck Profiles of Internet Buyers in 20 Countries: Evidence for Region-Specific Strategies. Journal of International Business Studies 32 (4): Lowe, A., A. Chun-Tung, and D. R. Corkindale Differences in Cultural Values and Their Effects on Responses to Marketing Stimuli: A Cross-Cultural Study Between Australians and Chinese from the People s Republic of China. European Journal of Marketing 32:

14 14 K. Goodrich and M. de Mooij Mediascope Europe EIAA Online Shoppers Accessed September 1, Magnusson, P., R. T. Wilson, S. Zdravkovic, J. X. Zhou, and S. A. Westjohn Breaking Through the Cultural Clutter: A Comparative Assessment of Multiple Cultural and Institutional Frameworks. International Marketing Review 25 (2): Nielsen Nielsen Global Survey Pornpitakpan, C Factors Associated with Opinion Seeking: A Cross-National Study. Journal of Global Marketing 17 (2/3): Schultz, D. E., and M. P. Block Antecedents and Consequences of Consumer Participation in Virtual Brand Communities in Social Networking. In The Proceedings of the 2009 American Academy of Advertising Asia-Pacific Conference, edited by Hairong Li, Shengmin Huang, and Dinghai Jin. Shanghai, China: American Academy of Advertising, in conjunction with China Association of Advertising of Commerce, and Communication University of China. Soares, A. M., M. Farhangmehr, and A. Shoham Hofstede s Dimensions of Culture in International Marketing Studies. Journal of Business Research 60: Søndergaard, M Research Note: Hofstede s Consequences: A Study of Reviews, Citations and Replications. Organization Studies 15: Specht, N How Social Media is Used in Germany, China and Brazil. Accessed May 6. Marketers.aspx Su, N., Y. Wang, G. Mark, T. Aiyelokun, and T. Nakano A Bosom Buddy Afar Brings a Distant Land Near: Are Bloggers a Global Community? Proceedings of the Second International Conference on Communities and Technologies. Milano, Italy, June Unesco Institute for Statistics Newspapers Per 1,000 People. Accessed September 3, United Nations Development Program Report Adult Literacy Rates. Accessed August 3, Van Belleghem, S Social Media Around the World. InSites Consulting. Accessed February 22, Vasalou, A., A. N. Joinson, and D. Courvoisier Cultural Differences, Experience with Social Networks and the Nature of True Commitment in Facebook. International Journal of Human Computer Studies 68: Westbrook, R. A Product/Consumption-Based Affective Responses and Postpurchase Processes. Journal of Marketing Research 24: World Values Survey Accessed August 5,